Re: [Boost-bugs] [Boost C++ Libraries] #6126: Signed integer members of Boost.Fusion adapted ADTs are not output correctly with Boost.Spirit.Karma rules

Subject: Re: [Boost-bugs] [Boost C++ Libraries] #6126: Signed integer members of Boost.Fusion adapted ADTs are not output correctly with Boost.Spirit.Karma rules
From: Boost C++ Libraries (noreply_at_[hidden])
Date: 2012-02-02 01:14:58

#6126: Signed integer members of Boost.Fusion adapted ADTs are not output
correctly with Boost.Spirit.Karma rules
-------------------------------+--------------------------------------------
  Reporter:  t0rt1e@…          |       Owner:  hkaiser                                                                                                             
      Type:  Bugs              |      Status:  new                                                                                                                 
 Milestone:  To Be Determined  |   Component:  spirit                                                                                                              
   Version:  Boost 1.48.0      |    Severity:  Regression                                                                                                          
Resolution:                    |    Keywords:  patch proposed, Boost.Spirit.Karma, Boost.Fusion, BOOST_FUSION_ADAPT_CLASS, BOOST_FUSION_ADAPT_ADT, short, int, long
-------------------------------+--------------------------------------------

Comment (by hkaiser):

 Replying to [comment:13 bugs@…]:
 > I tracked it down to
 extract_from_attribute<>::call<>(attr,ctx,true_type). It will generate an
 inner typedef of
 >
 >         typedef typename mpl::eval_if<
 >             is_one_element_sequence
 >           , detail::value_at_c<Attribute, 0>
 >           , mpl::identity<Attribute const&>
 >         >::type type;
 >
 > which is going to be a reference. This is ok for tuples, adapted structs
 (typical fusion sequences) but not for ADT's as the accessor function
 returns a `short`, which **can** be bound to a const& **but** only locally
 (its lifetime won't be extended outside the scope in which the reference
 is declared).
 >
 > Now the way I see it in the code, the sample from the docs[1] `(int,
 int, obj.get_age(), obj.set_age(val))` won't work without undefined
 behaviour, precisely because it results in
 `extract_from_attribute<>::call<>` returning a reference to a local (`int`
 in that case).
 >
 > For the OP, I guess the best way to fix it would be to make the
 accessors return (const) references: (see attached)
 >
 >     struct XR
 >     {
 >         short x;
 >
 >         explicit XR(short num = 0) : x(num) {}
 >         const short& getx() const { return x; }
 >               short& getx()       { return x; }
 >         void  setx(short v)       { x = v; }
 >     };
 >
 >     BOOST_FUSION_ADAPT_ADT(XR, (short&, const short&, obj.getx(),
 obj.setx(val)))
 >
 > This also removed all the other valgrind messages that were previously
 unaccounted for.
 >
 > [1]
 http://www.boost.org/doc/libs/1_48_0/libs/fusion/doc/html/fusion/adapted/adapt_adt.html


 That's excellent news! Thanks. That allows to revert the patch above.

-- 
Ticket URL: <https://svn.boost.org/trac/boost/ticket/6126#comment:15>
Boost C++ Libraries <http://www.boost.org/>
Boost provides free peer-reviewed portable C++ source libraries.